Adaptive iterative learning control of uncertain robotic systems

Abstract

A distinct feature of the proposed AILC scheme is that uncertain parameters are estimated in the time domain whereas repetitive disturbances are identified and compensated in the iteration domain. The bounds of the parameters are not required to be known a priori, and the learning control gain can be adjusted independently of the parameter adaptation gain. The overall closed-loop stability and uniform error convergence in the iteration domain are established without any acceleration measurements or their estimated values. The proposed AILC scheme is a balanced combination of the conventional adaptive control and the iterative learning control, where the shortcomings of each scheme are complemented. The validity of the scheme is confirmed through a simulation example.